# OOM run\_seq2seq.py from checkpoint

**URL:** </static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Foom-run-seq2seq-py-from-checkpoint%2F4241%3E%5C
**Category:** 🤗Transformers\
**Created:** [March 8, 2021, 2:28pm UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Foom-run-seq2seq-py-from-checkpoint%2F4241 "2021-03-08T14:28:04Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![smaximo](https://avatars.discourse-cdn.com/v4/letter/s/6f9a4e/32.png) [@smaximo](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Fu%2Fsmaximo)%5C
**Post date:** [March 8, 2021, 2:28pm UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Foom-run-seq2seq-py-from-checkpoint%2F4241%2F1 "2021-03-08T14:28:04Z")

</div>

If I fine-tune run\_seq2seq.py with a specific model from the hub and a specific dataset in 4 epochs it finishes correctly.  
I save a checkpoint at the third epoch.

But if I load the checkpoint at the third epoch and try to continue until the forth I get an out of memory error.

Is it necessary to make adjustments in the arguments of run\_seq2seq.py, when starting from a checkpoint, to obtain similar results?